Dynamic random access memories (DRAMs) are data storage devices that store data as charge on a storage capacitor. A DRAM typically includes an array of memory cells. Each memory cell includes a storage capacitor and an access transistor for transferring charge to and from the storage capacitor. Each memory cell is addressed by a word line and accessed by a bit line. The word line controls the access transistor such that the access transistor controllably couples and decouples the storage capacitor to and from the bit line for writing and reading data to and from the memory cell.
The storage capacitor must have a capacitance that is large enough to retain a charge sufficient to withstand the effects of parasitic capacitances, noise due to circuit operation, and access transistor reverse-bias junction leakage currents between periodic data refreshes. Such effects can result in erroneous data. Obtaining a large capacitance typically requires a storage capacitor having a large area. However, a major goal in DRAM design is to minimize the area of a DRAM memory cell to allow cells to be more densely packed on an integrated circuit die so that more data can be stored on smaller integrated circuits.
In achieving the goal of increasing DRAM array capacity by increasing cell density, the sufficient capacitance levels of the DRAM storage capacitors must be maintained. A "stacked storage cell" design can increase the cell density to some degree. In this technique, two or more capacitor conductive plate layers, such as polycrystalline silicon (polysilicon or poly), are deposited over a memory cell access transistor on a semiconductor wafer. A high dielectric constant material is sandwiched between these capacitor plate layers. Such a capacitor structure is known as a stacked capacitor cell (STC) because the storage capacitor plates are stacked on top of the access transistor. However, formation of stacked capacitors typically requires complicated process steps. Stacked capacitors also typically increase topographical features of the integrated circuit die, making subsequent lithography and processing, such as for interconnection formation, more difficult Alternatively, storage capacitors can be formed in deep trenches in the semiconductor substrate, but such trench storage capacitors also require additional process complexity. There is a need in the art to further increase memory storage density without adding process complexity or additional topography.
Electrically erasable and programmable read only memories (EEPROMs) provide nonvolatile data storage. EEPROM memory cells typically use field-effect transistors (FETs) having an electrically isolated (floating) gate that affects conduction between source and drain regions of the FET. A gate dielectric is interposed between the floating gate and an underlying channel region between source and drain regions. A control gate is provided adjacent to the floating gate, separated therefrom by an intergate dielectric.
In such memory cells, data is represented by charge stored on the polysilicon floating gates, such as by hot electron injection or Fowler-Nordheim tunneling during a write operation. Fowler-Nordheim tunneling is typically used to remove charge from the polysilicon floating gate during an erase operation. However, the relatively large electron affinity of the polysilicon floating gate presents a relatively large tunneling barrier energy at its interface with the underlying gate dielectric. The large tunneling barrier energy provides longer data retention times than realistically needed. For example, a data charge retention time at 85.degree. C. is estimated to be in millions of years for some floating gate memory devices. The large tunneling barrier energy also increases the voltages and time needed to store and remove charge to and from the polysilicon floating gate. "Flash" EEPROMs, which have an architecture that allows the simultaneous erasure of many floating gate transistor memory cells, require even longer erasure times to accomplish this simultaneous erasure. The large erasure voltages needed can result in hole injection into the gate dielectric. This can cause erratic overerasure, damage to the gate dielectric, and introduction of trapping states in the gate dielectric. The high electric fields hat result from the large erasure voltages can also result in reliability problems, leading to device failure. There is a need in the art to obtain floating gate transistors that allow the use of lower programming and erasure voltages and shorter programming and erasure times.